The overall goal of this research is the development of a comprehensive, non-invasive MRI based method to quantify blood flow in the hepatic vasculature of patients with portal hypertension. Portal hypertension is the most common and lethal complication of end-stage chronic liver disease and leads to the development of portosystemic collaterals (varices) that shunt blood from the liver, renal dysfunction (hepatorenal syndrome), accumulation of ascites, among other complications. Portal hypertension is currently assessed using invasive catheter-based measurements of hepatic venous pressure gradient (HVPG). HVPG is the best-validated biomarker of portal hypertension, but only provides a global measure of disease. Most therapies reduce portal pressure through reductions in the flow to the liver (eg. beta-blockers) or shunting flow from the portal circulation (TIPS procedure). There is need for accurate biomarkers of blood flow to the liver that would complement and augment HVPG measurements. However, there are currently no available methods to provide such hemodynamic information. This proposal seeks to develop a velocity sensitive 4D flow MRI method that simultaneously images the hepatic vascular anatomy and quantifies blood flow within all vascular territories of the liver. This approach is based on established and validated radially undersampled k-space trajectories (PC VIPR) and will provide a rapid image acquisition with advanced velocity encoding to improve velocity sensitivity, matching the range of flow velocities encountered in the liver. This acquisition will be coupled with non-Cartesian parallel imaging to further reduce scan time to approximately 10 min of free breathing while maintaining high spatial (<1.3 mm isotropic), and high temporal resolution (<50 ms) over a large imaging volume with good image quality and accurate flow quantification. The accuracy of this approach will be evaluated in the hepatic vasculature of swine using perivascular ultrasonic flow probes as the reference standard. The precision (repeatability) of the flow measurements will be determined in patients with portal hypertension and in normal subjects while also quantifying factors that impact precision (meals, diurnal variation). Finally, the clinical utility of the apprach will be determined with non-invasive flow measurements in the liver as part of a pilot study that uses a well- validated beta-blocker test-dose paradigm in patients with portal hypertension. The decrease in hepatic venous pressure gradient (HVPG) will be correlated with decreases in portal blood flow achieved using test dose of propranolol, a non-selective beta-blocker. This study will demonstrate the potential clinical utility of 4D flow MRI as a non-invasive surrogate biomarker of HVPG. If successful, 4D flow MRI would provide a comprehensive tool to quantify and characterize blood flow in portal hypertension, improving risk stratification for variceal hemorrhage and providing non- invasive surrogate biomarkers to evaluate the response of portal flow to pharmacological challenges.